Plasma and red cell sorbitol assay in diabetic subjects.

Plasma and red cell sorbitol concentrations, fasting plasma glucose, glycosylated hemoglobin (GHb) were evaluated in 30 diabetic patients and 42 normal subjects. Red cell sorbitol levels were evaluated in hemolysate and non-hemolyzed samples. Mean red cell sorbitol concentrations evaluated after hemolysis were 21.5 +/- 5 nmol/ml in diabetics and 12.2 +/- 4 nmol/ml in controls (p less than 0.001). Mean values of red cell sorbitol determined in non-hemolyzed samples were 13.9 +/- 3 nmol/ml in diabetics and 8.1 +/- 3 nmol/ml in controls (p less than 0.001). Mean plasma sorbitol concentrations were 8.4 +/- 3 nmol/ml in diabetics and 5.2 +/- 1 nmol/ml in controls (p less than 0.001). The within and between run reproducibilities evaluated in plasma, hemolysate and non-hemolyzed red cells gave the lowest values for hemolyzed red cell samples. The studies on stability showed that samples neutralized and stored at -20 degrees C gave reproducible values if assayed within 3 days. A positive correlation was found between red cell hemolysate sorbitol concentrations and fasting plasma levels in both diabetics (r = 0.56; p less than 0.001) and controls (r = 0.62; p less than 0.001). Red cell hemolysate sorbitol concentrations were also positively correlated to GHb in both diabetics (r = 0.64; p less than 0.001) and controls (r = 0.36; p less than 0.05). We believe the most effective sorbitol assay to be that obtained in hemolyzed red cells and that these values related to glycemic levels could be a useful index of metabolic control and an indicator of tissue sorbitol levels in humans.